Re: XyWrite Development

[OkAnnie@xxxxxxxx]

Kenneth Frank: Thanks for your attention to the UD and capped
mode issues. I regret finding myself in this contentious
position, but I'm going to try to detail what anyone with much
xpl code or highly customized .kbd files faces when upgrading
with TTG's documentation of function symbol redefinitions.

First, however: The unconditional v3 rubout ... UD functions do
*exactly* what I want when copying or moving text. Since my
earliest v3 days, CP and MV have been absent from my .kbd and xpl
files. They don't interest me as a subject for analysis, but
"Command Reference Guide" pp 3-25f Notes #2 seem to describe the
kind of restrictions I remember.

"Making the Transition" is where one would expect an explanation
of any v3->4 function symbol redefinitions. Of the two noted
there: "The RE (Rubout to End of Line) function call now moves
the cursor to the beginning of the next line if the cursor is at
the end of the line. In XyWrite III Plus, the cursor remained on
the same line."

The other change pertains to DW. UD is not mentioned.
"Customization Guide" says merely that UD restores the last text
deleted (this would be consistent with v3 UD), and is assigned to
 and . In the "Command Reference
Guide" index, "Undoing," "Undeleting," and "Undeleting Text" omit
references to function symbols. Among the nine citations, try
"Undeleting text." Learn on p 3-30 that--before redisplaying at
the cursor position what is on the stack-- produces a dialog
box showing the first 50-60 chars of X number of blocks of
recently deleted text; user highlights the text to be restored
and activates the "Play Back" or "Entire Block" "pushbutton."
Nice! But I shudder at what would happen if my countless xpl UDs
encountered a dialog box where they expect to find one block on
the stack and redisplay it.

Back to CG to see if the simple procedure UD did before has been reassigned.
Browse the terse function descriptions: New function UN, assigned to , pastes copy from clipboard. Back to the CRG index to find
"clipboard"; index lists no function symbol.  appears
under Copying and Moving a Block of Text only in notes that say
 restores blocks defined with  and  respectively. Could  and  be CP and MV, which CG says are on  and ?
Nowhere in CRG or CG is it told what functions are assigned to
 and . v3 UD restores what's been removed
with the rubout functions. Does v4 UN? Who can say?

Looking now at CRG p 3-31 with calmer eye than when I embarked on
the v4 treasure hunt, I see Note #5: "Setting the D1 default to 0
disables the undelete stack. ... XyWrite saves only the most
recently deleted text in memory. ... retrieve it by pressing
." Sounds like v3 UD! Does this means no dialog box? Do I
have to reinstall v4 to find out?

"Making the Transition" devotes slightly more than one loose
page, 14f, to function and command changes in terms of xpl. If the changes to UD aren't in
MTT why would I not expect to go through this grief for every
function symbol?

"Customization Guide" lists functions alphabetically with a few
descriptive words each and the symbol's key assignment.

"Command Reference Guide" describes procedures associated with
given keys but gives function symbol references only in the
index, if at all, and the index omits key assignments.

Nowhere is there a systematic presentation of function symbols by
key assignments--e.g., kbd maps of what function is assigned to
each key in each unshifted, shifted, and supershifted position.

For information beyond the terse the CG list, the user must note
the key assignment and browse the CRG index till finding an entry
for a procedure likely to be associated with the key.

Going the other way, to find what function symbol is associated
with a description in CRG, one must note the key assignment and
browse the CG list right side, which is disorderly since the list
is alphabetized by function.

Just as the CRG text always and the index often omits the
function symbol, the CG list doesn't list all key assignments.
...

M0-M9 functions are on Alt keys in my .kbds (save/gets are the
Ctrl+Shift set) but I never use them. I can't explain why; they
just aren't right for the way I work. Almost from the start, I've
done attributes another way.
Typically, my  (=roman) cancels other styles this way:
   19=SI,{,m,d,n,m,},CI
Since no XP/TP pair is present this should cause a split-second
"Invalid formatting command" error; if it happens it's too fast to see even on an XT.
This works for me, and I suspect that it's what xyDos 4 rebelled against. The
.kbd fix is easy and obvious. But my copious xpl files change
attributes with save/gets, always lc, throughout. What does this
new restriction gain? I threw up my hands. Out with the xy4!

This via email from another subscriber after I mentioned the, uh,
feature: "I will tell you, BTW, that they have _not_ solved the
lower case mode code problem. As you know, in 3+ if you enter
mdul in expanded mode, then go to normal mode and return to
expanded, you will find MDUL. In 4, not so. When it's in the
mood, it _will_ find mdul if you search for it; if not, not."
   --Annie

========================== annie fisher  nyc